A review of empathy education in nursing
Scott Brunero, Mental Health Liaison Nursing, Prince of Wales Hospital, EBB, NERU, rm 7, High St, Randwick, 2031 NSW, Australia.
E-mail: <firstname.lastname@example.org>; <email@example.com>
BRUNERO S, LAMONT S and COATES M. Nursing Inquiry 2010; 17: 65–74
A review of empathy education in nursing
The ability for nurse educators to improve the empathy skill set of nurses has been the subject of several studies with varied outcomes. The aim of this paper is to review the evidence for empathy education programmes in nursing and make recommendations for future nurse education. A review of CINAHL, Medline, Psych Info and Google Scholar was undertaken using the keywords empathy, person centredness, patient centredness, client centredness, education and nursing. The studies included were required to have measured the effectiveness of empathy training in postgraduate and or undergraduate nurses. The included studies incorporated both qualitative and quantitative methods and were published in peer-reviewed journals. Studies were ranked for level of evidence according to The Joanna Briggs Institute criteria. Seventeen studies from the literature review were found that met the inclusion criteria. Of the 17 studies, 11 reported statistically significant improvements in empathy scores versus six studies that did not. Several variables may affect empathy education that need to be accounted in future studies such as; gender, cultural values and clinical speciality experience. Models of education that show most promise are those that use experiential styles of learning. The studies reviewed demonstrated that it is possible to increase nurses’ empathic ability.
Engaging patients is seen as a critical part of the nurse–patient relationship with empathy being reported as an integral component of the relationship (McCabe 2004; Stein-Parbury 2005; Innes, Macpherson, and McCabe 2006). Reynolds, Scott, and Jessiman (1999) argued that the presence of empathy is critical to the development of the therapeutic relationship, while Kalish (1971) suggests that empathy was the most important component of such a relationship. The use of empathy has been well documented across the health professions as a means of engaging patients; by nurses (Reynolds and Scott 2000), medical staff (Kim, Kaplowitz, and Johnston 2004), pharmacists (Lonie et al. 2005) and dentists (Molen, Klaver, and Duyx 2004).
Empathy has been proposed to be the ability to perceive the meaning and feelings of another and to communicate those feelings to the other person (Stein-Parbury 2005). Empathy as a therapeutic tool has its origins in the work of Carl Rogers (1959) who saw empathy as a core of his person-centred approach to counselling. Rogers (1959) describes empathy as: the state of perceiving the internal frame of reference of another person, with accuracy and with emotional components and meanings that pertain to it, as if one were with the other person, but without the loss of the as-if condition (Rogers 1959).
Barret-Lennard (1981) suggests that empathy is a cyclic model involving the following stages; understanding and recognising the other person (empathee’s) emotions, communicating this understanding to the empathee and recognising that this has been understood. Hojat et al. (2002) views empathy in two parts: cognitive empathy and affective empathy. The cognitive domain involves the ability to understand another person’s inner experience and feelings, with an ability to view the outside world from the other person’s perspective. The affective domain involves the entering into or joining in the emotional experience of the other, which may also define sympathy (Hojat et al. 2002). While it maybe difficult for health professionals to function solely in one domain, the health professional who acts in the affective domain could lose their objectivity and become overwhelmed by the emotions of the patients in their care. The affective domain of empathy has also been described as a maladaptive human response (Hojat et al. 2002).
Several previous studies have attempted to demonstrate the effectiveness of the use of empathy within health-care. Reynolds and Scott (2000) report a positive relationship between empathy and patient responses such as; relief from pain, improved pulse and respiratory rates, and clients self-report of worry and distress. Williams (1979) showed that nurses, who displayed high levels of empathy to institutionalised elderly patients, found that these patients experienced a statistically significant improvement of self-concept, as understood by a reduction in dehumanisation and depersonalisation. LaMonica et al. (1987) found less anxiety, depression and hostility in cancer patients being cared for by nurses who show high levels of empathy. Reynolds and Scott (2000) report that the quality of client’s self-disclosure was found to be associated with the level of empathy used by nurses. Efforts to use empathy skills, understand their influence on patient care and the ability of nurse’s to apply these skills would appear warranted.
Given that the aim of the study was to review the outcomes of empathy programmes in nursing and make recommendations for future nurse education, a review of CINAHL, Medline, Psych Info and Google Scholar was undertaken using the keywords empathy, person centredness, patient centredness, client centredness, training, education and nursing. Included studies were required to measure the effectiveness of empathy levels in postgraduate nurses and or undergraduate nurses, be either qualitative or quantitative in their research design and had been published in a peer-reviewed journal. Excluded studies were that of other professional disciplines. The time period for the publication of the studies was left open as the volume of papers is limited. The studies included in the review were analysed using a method suggested by The Joanna Briggs Institute (2007). Articles that contained quantitative data were ranked in terms of levels of evidence, with level 1 being evidence obtained from reviews of all randomised control trials (RCTs), level 2 being obtained from at least one properly designed RCT, level 3 (I) being obtained from a well designed control trial without randomisation, level 3 (II) being obtained from a comparative study without randomisation but with control and allocation, level 3 (III) being obtained from a comparative study with the control being historical and level 4 being obtained from case series, audits, questionnaires, surveys and literature reviews (The Joanna Briggs Institute 2007). In an attempt to understand the educational methods used within the studies, all the studies were analysed for the learning style described within the training programmes as defined by Laschinger and Boss (1984) and Burnard (1992).
Seventeen studies from the literature review were found that met the inclusion criteria. Two populations of nurses have been primarily studied, undergraduate nurses (table 1) and postgraduate nurses (table 2). One study combined both students and postgraduate nurses (Taylor et al. 2009). These study results are included in with the postgraduate nurses (table 2), as the authors identified difficulty in determining how many were students and postgraduates. According to The Joanna Briggs Institute’s (2007) evidence-level ratings of studies, two studies met criteria for level 2, seven for level 3(II), seven for level 3(III) and one for level 4. The researchers overall felt a need to study empathy training quantitatively. A total of nine empathy measurement tools that have had some level of validation were used in the studies. A range of qualitative measures such as participant’s satisfaction questionnaires and checklists were used to evaluate the studies.
Table 1. Undergraduates studies
|Cinar and Cevahir (2007)||3(III)||Undergraduates nurses|
n = 139
|Empathic skill scale||4-year undergraduate course|
|Empathy significantly increased between second and fourth year|
|Cutcliffe and Cassedy (1999)||3(III)||Quasi experimental|
n = 38
|Ivey empathy rating scale||12-day short skills-based course|
|Statistically significant improvement in empathy scores|
|Hodges (1991)||2||Undergraduates nurses, general nurses|
n = 13 empathy education versus n = 13 psychological mindedness education
|Video tape of students’ post education, rated by the patient interview assessment schedule||13 hours over 3 weeks, for empathy education and psychological mindedness education|
|No significant difference between groups|
|Nardi (1990)||3(III)||Post test control group design|
n = 35 student nurses
|Gazda empathy scale||3 hours of education, 2 hours for testing|
|Statistically significant improvement for intervention group|
|Rogers (1986)||3(III)||Undergraduates nurses n = 135 females||Empathy construct rating scale||Baccalaureate undergraduate nursing programme|
|Evans et al. (1998)||3(III)||Pre–post test undergraduate Bachelor of Nursing course students|
n = 10
|Layton empathy tests|
|3-year nursing degree|
|Empathy skills learnt in undergraduate course not sustained|
|Wikstrom (2001)||4||n = 428, first-year undergraduate nurses|
Sample selected randomly from among eight university colleges
|Critical discussion group||Visual art programme to increase empathy (small group work). Students studied Edward Munch’s The sick child and interpreted with a focus on personal knowledge of empathy|
|85% of students reported on themes of empathy|
Visual art programme stimulated students to discuss and write about empathy
Table 2. Postgraduate studies
|Wallston et al. (1978)||3(II)||Controlled trial|
n = 24 controls
n = 20 intervention
|Nurse responses audio taped transcribed and analysed for difference between groups||450-word statement, containing elements of a helpful response, with illustrative examples of empathy|
|Intervention group showed significant increase in person-centredness|
|Ancel (2006)||3(III)||Pre–post test|
n = 263 nurses
|Empathic communication skill B, participants satisfaction form||5 days of education|
|Statistically significant improvement for intervention group|
|Razavi et al. 2002||2||Pre–post, longitudinal and randomised|
n = 115 oncology nurses
n = 57 intervention
n = 58 control
|Reading of a role play with an actor, rating of words used||105 hours of education follow up at 3 and 6 months|
|Trend upwards in use of emotionally laden words|
|LaMonica et al. (1987)||3(II)||Evaluation of patients after nurse education on empathy skills|
n = 56 nurses intervention
n = 53 controls
n = 656 patients
|Multiple affect adjective checklist|
La Monica/Oberst patient satisfaction scale
Empathy construct rating scale
|14- to 16-hour programme used, topics helping model, empathic responses, communication theory, perception of verbal and non-verbal feelings, ineffective communication styles, care of oneself|
|Clients cared for by nurse trained group showed significantly less hostility and anxiety than controls|
|LaMonica et al. (1976)||3(II)||Pre–post test controlled study|
2n = 39 registered nurses
|Carkuff index of communication|
Carkhuff Empathy Scale
|Human relations model, didactic and experimental learning, 11 hours in total, spread over seven sessions|
|Education significantly raised staff empathy levels|
|Edwards, Peterson, and Davies (2006)||3(II)||Matched pair, pre–post test study|
n = 22 (women health nurses)
Randomly selected nurses, case scenarios read aloud nurses responded verbally and then were rated on active listening, assertiveness skills, etc
|Communications skills were assessed after case scenario read aloud using a communication skills checklist||90 minutes educational sessions over 12 weeks weekly or bi weekly case study problem-based approach, then self-directed reading|
Multiple component intervention best practice guideline
|Statically significant improvement in the quality of active listening, initiating statements, and frequency of initiating skills|
|Fatma (2001)||3(II)||Quasi experimental design|
n = 43 nurses intervention
n = 70 controls
|Dokmens scale of empathic skills|
Empathic tendency scale
|20 hours of empathy skills education|
|Empathic communication skills developed in the intervention group, education did not impact on empathic tendency levels|
|Herbek and Yammarino 1990||3(II)||Controlled trial|
Medical surgical nurses
n = 16 intervention
n = 19 controls
|Mehrabian and Epstein emotional|
Empathic tendency scale
|6 hours of intensive education over 7 weeks|
|Statistically significant difference|
|Yates, Clinton, and Gary (1998)||3(III)||Postgraduate palliative care nurses|
Pre and post test
n = 181
|Staff–patient interaction response scale||14-week group session of 90 minutes each and peer consultation|
|Statistically significant improvement in empathic ability|
|Taylor et al. (2009)||3(II)||Pre and post|
n = 201
There was difficulty in determine the difference between students and RNs, which has been noted in the study
|Response empathy scale||Self-directed learning package, participants received a workbook and DVD, which contained the pre and post measures which were returned by mail. Ten hours was the suggested time to complete the package|
|Statistically significant improvement in empathic ability|
The majority of the papers (n = 10) were published before 2000, with seven being published after this. While empathy is not a new concept to nursing the more recent publications do support the currency of empathy education in the modern nursing environment. Most of the studies were completed in the UK or North America, with the study completed by Fatma (2001) in Turkey. Cultural variations on empathy were not discussed within the studies, yet nurses work within a wide range of diverse cultural and language groups.
Clinical specialties of the nurses within the studies included; four generalist nursing studies, one palliative care nursing study, two medical/surgical nursing studies, one oncology study and one women’s health nursing study. While a broad range of nursing specialties are described, little is discussed in the literature on the varying levels of empathic nursing skills needed for the nursing specialities (Cutcliffe and Cassedy 1999).
The interventions described ranged significantly in length. Two of the studies (Evans et al. 1998; Cinar and Cevahir 2007) considered the whole undergraduate training course as the intervention. Within the postgraduate education programmes, most programmes ranged from 6 to 105 hours in length, with one education programme consisting of a short reading task (Wallston et al. 1978). Statistically significant results were evident from the shorter duration studies 11 hours (LaMonica et al. 1976), with the longest programme of 105 hours showing only trend improvement (Razavi et al. 2002). Length of education is a significant variable as there are high costs associated with delivering education programmes to nurses. Of the 17 studies 11 described using experiential learning styles within their training programmes, with five using didactic styles of learning and one study using a self-directed learning package (tables 1 and 2, intervention column for each study learning style) (Laschinger and Boss 1984; Burnard 1992). Of the 11 studies that used experiential style learning, eight are reported to have statistically significant positive results on empathy skill improvement (tables 1 and 2, intervention and outcome column).
Of the 17 studies, 11 reported improvements in empathy skill level of participants. In the higher level evidence studies (level 2), one study showed trend improvement (Hodges 1991), with the other not showing any statistical significant improvement (Razavi et al. 2002). The small sample size in the Hodges’ (1991) study (n = 13, intervention group of undergraduates) would limit the generalisability of the results. Razavi et al. (2002) had a much larger sample (n = 57 intervention and n = 58 control) but chose unconventional measurement methods. This study did not use previously published validated measurement tools of empathy, instead asking their participants to rate empathic words after reading of a role play. Although these two studies attempted to use higher methodological approaches, they both are limited by their sample size and choice of measurement tool. The levels 3(II) and 3(III) studies showed with all but one of the studies, significant improvements on empathy scores. Sample sizes in these studies ranged from n = 10 to n = 263, with the non-significant study by Rogers (1986) having a large sample of n = 134. Overall sample sizes varied greatly across the all studies from n = 10 to n = 428.
One of the difficulties with the Rogers’ (1986) study was the length of intervention they considered, the whole 3-year undergraduate programme. The levels 3(II) and 3(III) studies mostly used validated measurement tools, thus enhancing their chance of demonstrating positive outcomes. The use of valid and reliable measurement tools are seen as critical part of constructing effective research methodologies (Burns and Grove 2002).
Overall agreement between researchers regarding the studies, which showed significant improvement, were primarily characterised by their shorter length of educational programme and use of a validated measurement tool. The vast majority of the study samples were dominated by females, which is typical within nursing studies given the reported gender bias in nursing research (Polit and Beck 2008).
The results yielded a range of measurement tools used which highlight the need to develop standardised ways to evaluate changes in empathy as a result of education (Reynolds and Scott 2000; Ancel 2006; Yu and Kirk 2008). Yu and Kirk (2008) reviewed measurement tools of empathy (n = 20) and concluded that there is no consistent use of measurement tools and a strong need to evaluate the rigour of instruments used. Tools that have been developed may not take into account patient views on empathy, or only measure particular components of empathy (Ancel 2006). The range of measurement instruments in the studies reviewed show only two instruments being used: the empathy construct rating scale (LaMonica et al. 1987) and the empathic tendency scale (Epstein 1972).
Understanding the time it takes to impart empathy skills to nurses is critical as there is an obvious cost to education. Several studies attempted to evaluate change in empathy scores in undergraduate education over a 3- to 4-year period. Many variables during the undergraduate nurse’s education could impact on their empathy skill level. As an example, the level of clinical exposure and number of clinical placements in areas of nursing where the nurse–patient relationship plays a significant part in care, such as mental health, could impact the level of nurse’s skill (Rogers 1986; Evans et al. 1998; Cinar and Cevahir 2007).
There were two studies that showed negative results, Hodges (1991) and Rogers (1986). Hodges (1991) attributes this to the type of intervention applied and the lack of reliability and validity of the instrument chosen to measure the change. Rogers (1986) reports the lack of measured change in a baccalaureate programme, because of the difficulty in separating out the empathy specific parts of education compared to the overall education programme. Studies completed over long durations have been reported elsewhere to be limited as there are variables that are out of control of the researcher (Burns and Grove 2002).
Empathic responses can be influenced by a number of variables; personality, gender, interpersonal style, culture, social confidence, environment and the level of communication skills that have been learnt (Alligood and May 2000). Empathy researchers will need to develop a greater understanding of these variables to further the ability of education to improve empathic response. Chung and Bernak (2002) discuss the importance of culture and empathy within a counselling framework. Suggesting the key to being culturally empathic is to accept the culture, values and beliefs of others, while retaining your own culturally identity. There was limited attempt to control for these variables within the studies reviewed. None of the studies tried to determine any differences between genders. A probable reason for this is simply the limited number of males within the nursing profession making sample sizes of males within the studies too small for analysis. Our ability to empathise with others is argued to be gender influenced with reports of males demonstrating less empathic ability than females (Baron-Cohen 2002; Lawson, Baron-Cohen, and Wheelwright 2004).
The use of experiential learning styles was reported in the majority of the studies’ training programmes. This clear trend towards its use in empathy training is worthy of pursuing in further studies. Several of the studies describe using role play and case scenario-based experiential work (Tables 1 and 2, noted in intervention column as ‘experiential’). Case scenario-based learning is a subset of experiential learning, including aspects of problem-based learning and simulation, and involves using cases for teaching and getting students to problem solve clinical scenarios (Delpier 2006). Of the studies reviewed, Edwards, Peterson, and Davies (2006) used a case study-experiential learning approach, reporting statistically significant improvements in nurse’s empathic ability. Its benefits are said to allow students to test out a variety of clinical styles or judgements with the minimum of risk (Delpier 2006). Clinical cases can use realistic content and events to create scenarios that accurately reflect the clinical setting (Krautschield, Kaakinen, and Rains-Warner 2008). The use of role play within clinical scenarios allows participants to explore a variety of roles that may be different to their own. When designing empathy education, educators need to place the nurse in the role of patient and the nurse, as to give the nurse the opportunity to reflect and understand the patient’s emotional state in a controlled situation (Colier 1999). Reflections on the role play experience by the participants, is where new concept formation and learning is achieved (Colier 1999). This may be more didactic, allowing teaching of the advanced communication skills necessary to become empathic. This then leads to the participant generating new ways of working, incorporating new skills into their current practice and reflecting on how it may be useful. Educators should link the reflection to the new concepts of how empathic behaviour can reduce workloads, including more accurate patient assessment and reduction in patient hostility (Reynolds and Scott 2000).
Without the experiential aspect of learning, there is a concern that skills are only taken on superficially, where knowledge may increase but behaviour change does not. Ramsden (1992) argues that experiential learning styles can countenance this and generate the deep learning that is required. Experiential learning needs to revolve around the day-to-day work of the nurse and reflect the high demands that are placed on nurses (Hamilton 2008).
Future implications for empathy education for nurses
Future empathy education needs to consider activities to enhance self-awareness and assessment of the clinician’s personal value systems (Halpern 2003). The suggestion here is that one must increase the understanding of their own values, before one can attempt to understand another’s (Reynolds, Scott, and Jessiman 1999; Stein-Parbury 2005). Stein-Parbury (2005) suggests that identifying accurately with the patient moves the helping relationship to a more intimate level. As such, being able to distinguish when and where to use empathy is important as it may be perceived as an invasion if poorly judged. James (1989) describes emotional burden as being the work generated by dealing with other people’s feelings. It is argued that being empathic in an unskilled way can lead to over involvement and in some cased increase depression (Hojat et al. 2002). The role of managing ones own emotional response when being empathic needs to be highlighted in education programmes. Processes such as clinical supervision have demonstrated effectiveness on reducing the stress and the emotional labour of nursing (Brunero and Stein-Parbury 2007).
There was no reflection on the issue of state versus trait empathy within the studies reviewed. Empathy researchers could identify study participants empathy skills scores by using, for example, the Empathy Construct Rating Scale (LaMonica 1981), which can determine trait empathy. Individuals with low scores on trait empathy could be identified for more in-depth training and development. Conversely, exposing people with high-trait empathy to full training may be unnecessary and ultimately lead to a higher cost in training. Kunyk and Olson (2001) concept analysis found five conceptualisation of empathy in nursing; (i) as a human trait, (ii) a professional state, (iii) a communication process, (iv) a caring process and (v) as a special relationship. In Kunyk and Olson’s (2001) professional state; empathy can be learnt and uses cognitive and behavioural components to express understanding of the patient’s reality back to them. The tension in this conceptual model is one of state versus trait empathy. Alligood and May (2000) argue that empathy is based on inherent personal ability. Morse and Pooler (2002) suggest that learned empathy is a second-level empathic response, suggesting the learned response keeps the caregiver somewhat detached, objective and therapeutically at arms length. Hojat et al. (2002) also support this view. Morse and Pooler (2002) argue that state-level empathy is learnt and therefore open to change, while Alligood and May (2000) argue that trait levels of empathy are difficult to modify. Further studies that can determine this difference will lead to more effective training programmes.
Our study was limited by the use of English language publications and therefore our understanding of empathy in a western nursing cultural context. Nursing across the world is highly mobile and needs to survive in a multicultural context; the use of empathy will need to be examined within this context in future studies.
Efforts to determine the effectiveness of empathy-based education have shown positive results, although in varied populations of nurses using a range of measurement tools. For empathy education to evolve, its effectiveness will need to be seen in well-designed trials, which demonstrate not only its effectiveness but also the pragmatic issues of delivering education in both the undergraduate and postgraduate nurse populations. Generalisability of the results is difficult due to the number of measurement tools and different styles of education used.
Being empathic has a significant place in nursing within the therapeutic relationship. The studies reviewed demonstrated that it is possible to increase nurse’s empathic ability from a range of clinical specialties and at both undergraduate and postgraduate levels.